LuxCoreRender Forums

Hi, I'm trying to make sphere lights work, but it looks like MIS doesn't work on them. In the attached image, there are 2 lights. Left is sphere light, right is mesh with emission (which luxcore logs as "light source with 1280 triangles"). (right side of image is denoised)

On sphere light, it looks like only light samples are taken into account, and with mesh lights with higher triangle count light sampling doesn't work (problem with smaller lights).

So the question is, is there any way to make this work? Is this a bug? (both light samples only on sphere light, and no light samples on higher poly lights)
I'm working with build from latest commit.

Bonus question: is there a way to make sphere light visible to camera?

The same problem is with "distant" light with large angle

Hi!

1. A quote about Point Lamp Radius property from BlendLuxCore LightPanel Point documentation.

Radius: Defines the size of the point lamp. If greater than zero, the point lamp is a sphere, but only in shadow and reflection calculations. It can not be viewed directly, so it will always be invisible to direct camera rays. A higher radius leads to softer shadows.

https://wiki.luxcorerender.org/BlendLux ... anel#Point

2. No.

Because there really is no "Sphere Light", but just a simple & effective computing model of an "infinitely small point light" - just an abstraction.
Sphere primitive should be supported first... but that's another topic. So you're left with triangles emitting light.
More about Physically Based Rendering.

valenn wrote: ↑Fri Jul 12, 2019 1:48 pm I'm trying to make sphere lights work, but it looks like MIS doesn't work on them. In the attached image, there are 2 lights. Left is sphere light, right is mesh with emission (which luxcore logs as "light source with 1280 triangles"). (right side of image is denoised)

There is no MIS because the point light with radius is not intersectable (as you can see, it is not visible) so it lacks the multiple way to sample the light, it is done only throw direct light sampling. And this leads to the result you are observing with a nearly specular glossy BRDF.

To have MIS, you need some kind of light source that is intersectable.